The Futility and Hazards of Boosterism

6 January 2022

by Professor David Livermore

The Government’s answer to Omicron is boosters, boosters and more boosters. Everyone over 18 is eligible. Early queues exceeded five hours; December 18th saw 904,000 boosted. Israel is doling out fourth shots to the over-60s and Germany plans to follow suit.

Please don’t lump me with the anti-vaxxers as I roll my eyes. The present vaccines are invaluable for preventing severe COVID-19 – probably via a T-cell mechanism – in anyone of late middle age or above, or with vulnerabilities.

But those of us who broadly support vaccination must be honest: Spring’s fond notion that we could mass vaccinate our way out of the pandemic, as with measles or polio, is a busted flush. The dash for boosters only underscores the point.

The core problem is that the vaccines give only brief protection against infection. Moreover, they are ‘leaky’, and SARS-CoV2 can outflank them. Clinical trials suggested that the doubly vaccinated initially enjoyed 90% protection against infection. But, with ageing vaccination and circulating Delta, this eroded by summer’s end. Omicron is even more evasive. At a December party in Oslo, where 98% of 117 guests had been double vaccinated, 74% caught COVID. Boosters restore protection to 71-76%, but this doesn’t last long. Early UKHSA evidence is that it’s down by 15-25% within 10 weeks.

Worse, by using leaky vaccines, we are conducting a huge Darwinian experiment, with ourselves as guinea pigs. Over many years, working on new antibiotics, I’d perform serial passage’ tests. These involve stressing bacteria by growing them with the antibiotic, then progressively upping its concentration. If resistance emerged easily, it’s likely that the antibiotic has a problem, and there is a guilty “Told you so” satisfaction – I admit – to seeing the same mutations in treatment failures as we’d first selected in the lab. Covid now looks eerily similar. Leaky vaccines will put a drag on replication of the classical virus but less so on evasive variants. These will proliferate, coming to dominate in the pool of virus.

The origins of Omicron itself are uncertain. It is remote in sequence from previously prevalent variants. Perhaps it evolved in Darkest Africa, where there is little testing; perhaps in a chronically-infected immunosuppressed patient, or perhaps via transmission to and from an animal host. One study suggests that it is adapted to mice, and it’s teasingly possible that these pesky rodents find discarded face-masks useful as nest-liners.

Whatever its origin, serendipity has put Omicron where we want it – causing milder disease. We should thank our lucky stars, not stress it to evolve further. The U.K. authorities initially dismissed this mildness – suggesting that it just reflected so many South Africans having been infected previously. But the mildness is now indisputable: the massive mid-December rise in U.K. cases, including among London’s over-60s, has not been followed by a commensurate rise in deaths, even three weeks later. Hospital admissions are rising but are notoriously being inflated by inclusion of those infected while in hospital, those admitted for other reasons, and by outbreaks in mental health trusts. An unpublished study from Hong Kong indicates that Omicron proliferates in the upper airways, but less in the lung, providing an explanation for reduced severity. This view is now supported by a separate U.K. study.

Given its mildness, Omicron should be the best foundation for a lifetime’s equilibrium with SARS-CoV-2, which is what we all face. Whereas vaccine-based immunity is brief and entails only a response to the viral spike protein, post-infection immunity is broader, targeting multiple components. This should make escape harder. Two large Israeli studies – here and here – and one from Denmark show that post-infection immunity persists longer than vaccine induced.

Some will say ‘We can keep suppressing emerging variants with tweaked vaccines’. I thought this once, but recant. Aside from logistic impossibilities – Omicron is spreading far faster than tweaked vaccines could be deployed – there’s the troublesome matter of ‘Original Antigenic Sin’. Essentially, our immune system is lazy with repetitive tasks. Once trained to respond to a virus, or fragment, it assumes, on seeing something similar, that it’s a reprise of the first challenge. If you seek to cover multiple variants you should vaccinate with a mix at the start, not sequentially. That can’t be done in a running fight.

There is one last aspect, flagged regularly here. It’s that for many weeks before the start of boosters (now distorting the pattern) the UKHSA reported higher ‘apparent’ infection rates in the vaccinated than in the unvaccinated for multiple age cohorts, even though hospitalisations and deaths remained lower in the vaccinated. This led to the ONS accusing the Agency of encouraging anti-vaxxers. Consequently, the anomalous data, once shown as histograms, are now relegated to tables with copious caveats.

Favoured explanations are that incorrect population denominators underestimate the infection rate in the unvaccinated, or that there are behavioural differences between the vaccinated and the unvaccinated. Maybe so. But an alternative explanation is that Geert vanden Bossche was right to fear that leaky vaccines would deliver a brief adaptive protection against infection but simultaneously impair innate immunity, causing a protracted reduction in non-specific protection once the initial protection wore off. Although I’m underwhelmed by molecular evidence for this hypothesis, it would accommodate the UKHSA data, and needs to be refuted.

Time as a bench scientist taught me not to airbrush unexpected and unwelcome results. So often they’re telling you something. I spent two postdoctoral years struggling to make a standard permeability assay work for a bacterium, Pseudomonas aeruginosa, where accepted wisdom was that antibiotic resistance reflected (im)permeability. Regardless of how I tweaked the assay my experiments refused to confirm this. At first I didn’t believe my results; then, slowly, I did. Next, I struggled with referees who suspected my lousy lab technique. Eventually, I managed to publish, saying that there must be another factor in play, though I didn’t know what. Elucidation came from Hiroshi Nikaido at Berkeley, with me as a bit-part collaborator, showing that resistance was a correlate of the Pseudomonas’s unexpected ability to pump antibiotics back out, not of restricting their entry. Efflux, not impermeability. We struggled to publish that, too, with grief from Hiroshi’s former PhD student, but now it’s accepted dogma.

These experiences taught me to ask ‘What do these data show?’, not to mould them round my presuppositions. And if vanden Bossche is correct we’ll face a worse Omicron spike among the vaccinated by spring, as the booster’s immediate protection wanes, along with redoubled undesirable evolutionary stress on the virus.

Adding everything together there remains a valid argument for boosting those likely to develop severe or life-threatening Covid, meaning the over-50s and anyone with specific risk factors. I have repeatedly urged these folks to be vaccinated, and I do so here too, though I’ll rile some readers. However imperfect, the vaccines remain their best line of defence if you’re at real risk.

But the argument for boosting the young and healthy with leaky vaccines is thinner than threadbare. At best it will give brief protection. At worst, it will misalign their immune systems and push the virus to evolve further and unpredictably.

As I write, there is one hopeful sign: Sir Andrew Pollard of the JCVI has said: “We can’t vaccinate the planet every six months.” Quite. And this time, unlike with the unwarranted vaccination of children, the JCVI must stick to its guns.

Dr. David Livermore is Professor of Medical Microbiology at the University of East Anglia.